Strange behavior of serial console under 2.4.9

Strange behavior of serial console under 2.4.9

[H . J . Lu]

The serial console under 2.4.9 is very strange. It is very slow. I have
no such problem with 2.4.3/2.4.5. Telnet is fine.


H.J.

Re: Strange behavior of serial console under 2.4.9

[Bradley D. LaRonde]

I haven't noticed that.  I just ran top with 0 delay at 115200 and it seems
normally fast to me.

Regards,
Brad

----- Original Message -----
From: "H . J . Lu" <hjl@lucon.org>
To: <linux-mips@oss.sgi.com>
Sent: Thursday, October 18, 2001 9:57 PM
Subject: Strange behavior of serial console under 2.4.9


> The serial console under 2.4.9 is very strange. It is very slow. I have
> no such problem with 2.4.3/2.4.5. Telnet is fine.

Re: Strange behavior of serial console under 2.4.9

[Bradley D. LaRonde]

I haven't noticed that.  I just ran top with 0 delay at 115200 and it seems
normally fast to me.

Regards,
Brad

----- Original Message -----
From: "H . J . Lu" <hjl@lucon.org>
To: <linux-mips@oss.sgi.com>
Sent: Thursday, October 18, 2001 9:57 PM
Subject: Strange behavior of serial console under 2.4.9


> The serial console under 2.4.9 is very strange. It is very slow. I have
> no such problem with 2.4.3/2.4.5. Telnet is fine.

Re: Strange behavior of serial console under 2.4.9

[H . J . Lu]

On Thu, Oct 18, 2001 at 10:11:18PM -0400, Bradley D. LaRonde wrote:
> I haven't noticed that.  I just ran top with 0 delay at 115200 and it seems
> normally fast to me.
> 

I am using 9600 buad. It used to be ok under 2.4.3/2.4.5. But under
2.4.9, the first 10 minutes after boot is very slow. After that, it
seems ok.


H.J.

Re: Strange behavior of serial console under 2.4.9

[Maciej W. Rozycki]

On Thu, 18 Oct 2001, H . J . Lu wrote:

> I am using 9600 buad. It used to be ok under 2.4.3/2.4.5. But under
> 2.4.9, the first 10 minutes after boot is very slow. After that, it
> seems ok.

 That might be driver-specific.  I'm using drivers/tc/zs.c and it works
fine at 115200 bps.

-- 
+  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
+--------------------------------------------------------------+
+        e-mail: macro@ds2.pg.gda.pl, PGP key available        +

Re: Strange behavior of serial console under 2.4.9

[Bradley D. LaRonde]

H.J., which serial driver are you using?

Regards,
Brad

----- Original Message ----- 
From: "Maciej W. Rozycki" <macro@ds2.pg.gda.pl>
To: "H . J . Lu" <hjl@lucon.org>
Cc: <linux-mips@oss.sgi.com>
Sent: Friday, October 19, 2001 9:26 AM
Subject: Re: Strange behavior of serial console under 2.4.9


> On Thu, 18 Oct 2001, H . J . Lu wrote:
> 
> > I am using 9600 buad. It used to be ok under 2.4.3/2.4.5. But under
> > 2.4.9, the first 10 minutes after boot is very slow. After that, it
> > seems ok.
> 
>  That might be driver-specific.  I'm using drivers/tc/zs.c and it works
> fine at 115200 bps.
> 
> -- 
> +  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
> +--------------------------------------------------------------+
> +        e-mail: macro@ds2.pg.gda.pl, PGP key available        +
> 

Re: Strange behavior of serial console under 2.4.9

[Bradley D. LaRonde]

H.J., which serial driver are you using?

Regards,
Brad

----- Original Message ----- 
From: "Maciej W. Rozycki" <macro@ds2.pg.gda.pl>
To: "H . J . Lu" <hjl@lucon.org>
Cc: <linux-mips@oss.sgi.com>
Sent: Friday, October 19, 2001 9:26 AM
Subject: Re: Strange behavior of serial console under 2.4.9


> On Thu, 18 Oct 2001, H . J . Lu wrote:
> 
> > I am using 9600 buad. It used to be ok under 2.4.3/2.4.5. But under
> > 2.4.9, the first 10 minutes after boot is very slow. After that, it
> > seems ok.
> 
>  That might be driver-specific.  I'm using drivers/tc/zs.c and it works
> fine at 115200 bps.
> 
> -- 
> +  Maciej W. Rozycki, Technical University of Gdansk, Poland   +
> +--------------------------------------------------------------+
> +        e-mail: macro@ds2.pg.gda.pl, PGP key available        +
> 

Re: Strange behavior of serial console under 2.4.9

[Ralf Baechle]

On Fri, Oct 19, 2001 at 03:26:10PM +0200, Maciej W. Rozycki wrote:

> > I am using 9600 buad. It used to be ok under 2.4.3/2.4.5. But under
> > 2.4.9, the first 10 minutes after boot is very slow. After that, it
> > seems ok.
> 
>  That might be driver-specific.  I'm using drivers/tc/zs.c and it works
> fine at 115200 bps.

I haven't noticed this problem ever on Origins which use the standard
16550 driver.

  Ralf

Re: Strange behavior of serial console under 2.4.9

[Jun Sun]

"H . J . Lu" wrote:
> 
> The serial console under 2.4.9 is very strange. It is very slow. I have
> no such problem with 2.4.3/2.4.5. Telnet is fine.
> 

That is usually a symptom when the serial interrupts are not correctly
delivered.

Jun

Re: Strange behavior of serial console under 2.4.9

[Bradley D. LaRonde]

Let's try that again...

Agreed - I've seen it once where it spit 16 chars, long pause, 16 more
chars, etc.  This was due to the serial driver not getting interrupts.

Regards,
Brad

----- Original Message ----- 
From: "Jun Sun" <jsun@mvista.com>
To: "H . J . Lu" <hjl@lucon.org>
Cc: <linux-mips@oss.sgi.com>
Sent: Friday, October 19, 2001 12:53 PM
Subject: Re: Strange behavior of serial console under 2.4.9


> "H . J . Lu" wrote:
> > 
> > The serial console under 2.4.9 is very strange. It is very slow. I have
> > no such problem with 2.4.3/2.4.5. Telnet is fine.
> > 
> 
> That is usually a symptom when the serial interrupts are not correctly
> delivered.

Re: Strange behavior of serial console under 2.4.9

[Bradley D. LaRonde]

Let's try that again...

Agreed - I've seen it once where it spit 16 chars, long pause, 16 more
chars, etc.  This was due to the serial driver not getting interrupts.

Regards,
Brad

----- Original Message ----- 
From: "Jun Sun" <jsun@mvista.com>
To: "H . J . Lu" <hjl@lucon.org>
Cc: <linux-mips@oss.sgi.com>
Sent: Friday, October 19, 2001 12:53 PM
Subject: Re: Strange behavior of serial console under 2.4.9


> "H . J . Lu" wrote:
> > 
> > The serial console under 2.4.9 is very strange. It is very slow. I have
> > no such problem with 2.4.3/2.4.5. Telnet is fine.
> > 
> 
> That is usually a symptom when the serial interrupts are not correctly
> delivered.

Backspace on Virtual Console causes oops

[Wayne Gowcher]

Dear All,

I am having a problem with the backspace key on a
virtual terminal on a 2.4.2 kernel. If I hit backspace
when there is input at the command line - no problem
it deletes the character before the cursor.
But if I press backspace when there are no characters
at the command prompt, the kernel throws an oops. The
process causing the oops is Bash.
Note, on the Serial console, backspace works fine all
the time.

Has anyone ever seen anything like this ? If you have
how did you solve it ?

Does anyone know the program flow / program details of
what happens when backspace is pressed eg :

    Key In > Key passed to Routine X > Routine X
passes it to Routine Y etc > Bash receives key.

Or could someone point me to a doc that may explain
this ?

Thx

__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com

Re: Backspace on Virtual Console causes oops

[Gleb O. Raiko]

Wayne Gowcher wrote:
> 
> Dear All,
> 
> I am having a problem with the backspace key on a
> virtual terminal on a 2.4.2 kernel. If I hit backspace
> when there is input at the command line - no problem
> it deletes the character before the cursor.
> But if I press backspace when there are no characters
> at the command prompt, the kernel throws an oops. The
> process causing the oops is Bash.
> Note, on the Serial console, backspace works fine all
> the time.
> 
> Has anyone ever seen anything like this ? If you have
> how did you solve it ?
> 

Guess 1: You've got PC UARTs and drivers/char/serial.c
Guess 2: TAB TAB in bash causes the oops too.
Guess 3: You didn't change famous beep function in the driver.


Regards,
Gleb.

Re: Backspace on Virtual Console causes oops

[Pete Popov]

On Fri, 2001-10-26 at 09:12, Wayne Gowcher wrote:
> Dear All,
> 
> I am having a problem with the backspace key on a
> virtual terminal on a 2.4.2 kernel. If I hit backspace
> when there is input at the command line - no problem
> it deletes the character before the cursor.
> But if I press backspace when there are no characters
> at the command prompt, the kernel throws an oops. The
> process causing the oops is Bash.
> Note, on the Serial console, backspace works fine all
> the time.

I'll make a guess.  With no input at the line, when you hit backspace,
the shell is probably generating a beep.  Your kernel is not handling
that.  I don't remember exactly where the "beep code" was but you can
probably find it pretty quickly.

Pete

> 
> Has anyone ever seen anything like this ? If you have
> how did you solve it ?
> 
> Does anyone know the program flow / program details of
> what happens when backspace is pressed eg :
> 
>     Key In > Key passed to Routine X > Routine X
> passes it to Routine Y etc > Bash receives key.
> 
> Or could someone point me to a doc that may explain
> this ?
> 
> Thx
> 
> __________________________________________________
> Do You Yahoo!?
> Make a great connection at Yahoo! Personals.
> http://personals.yahoo.com

Re: Backspace on Virtual Console causes oops

[James Simmons]

> I'll make a guess.  With no input at the line, when you hit backspace,
> the shell is probably generating a beep.  Your kernel is not handling
> that.  I don't remember exactly where the "beep code" was but you can
> probably find it pretty quickly.

linux/drivers/char/vt.c.

It is the functions kd_nosound and _kd_mksound.

Re: Backspace on Virtual Console causes oops

[Wayne Gowcher]

Gleb,

> Guess 1: You've got PC UARTs and
> drivers/char/serial.c
> Guess 2: TAB TAB in bash causes the oops too.
> Guess 3: You didn't change famous beep function in
> the driver.

Correct on all 3 counts. Thanks for pointing the way.

Also thanks to Pete Popov and James Simmons. In
particular to James for telling me to look at vt.c,
functions kd_nosound and _kd_mksound. I changed the
ifdef to be aware of my board and now the problem is
solved.

To All,

I am grateful that my problem was solved quickly. But
before I posted to the list I did a search on
"Backspace oops linux", "Backspace crash linux" and a
few other combinations to see if this was possibly a
known problem. But turned nothing up.
Is there any other list / repository I should be
looking at to pick up on bugs like this ?

Any tips / info would be greatly appreciated.

Wayne

__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com

Re: Backspace on Virtual Console causes oops

[Gleb O. Raiko]

Wayne Gowcher wrote:
> 
> Gleb,
> 
> > Guess 1: You've got PC UARTs and
> > drivers/char/serial.c
> > Guess 2: TAB TAB in bash causes the oops too.
> > Guess 3: You didn't change famous beep function in
> > the driver.
> 
> Correct on all 3 counts. Thanks for pointing the way.

In fact, guess 1 is irrelevant. :-)

> 
> I am grateful that my problem was solved quickly. But
> before I posted to the list I did a search on
> "Backspace oops linux", "Backspace crash linux" and a
> few other combinations to see if this was possibly a
> known problem. But turned nothing up.
> Is there any other list / repository I should be
> looking at to pick up on bugs like this ?
> 
> Any tips / info would be greatly appreciated.

This is a famous problem, just look at vt.c, you'll see a lot of ifdefs
around sound routines. Just every porting engineer who encounter this
problem solved it himself (and added own ifdef to vt.c). In my case, I
just got fault epc, found the address in objdump -D output and look at
the sources. For me it seems easier, than posting a message to a list.
<nothing personal, really>

Regards,
Gleb.

Re: Backspace on Virtual Console causes oops

[James Simmons]

> This is a famous problem, just look at vt.c, you'll see a lot of ifdefs
> around sound routines. Just every porting engineer who encounter this
> problem solved it himself (and added own ifdef to vt.c). In my case, I
> just got fault epc, found the address in objdump -D output and look at
> the sources. For me it seems easier, than posting a message to a list.
> <nothing personal, really>

I plan to have this problem removed in 2.5.X when the console system is
migrated to the input api :-)

ld error " linking PIC files with non-PIC files "

[Wayne Gowcher]

I am trying to cross compile X using the redhat7.0
distribution from the sgi mips site as my base. Most
files compile OK, but they fail at the link stage with
the following error :

linking PIC files with non-PIC files.

I created my cross-compile library files in
/usr/mipsel-linux from the packages :
glibc-2.2.2-1.mipsel.rpm &
glibc-devel-2.2.2-1.mipsel.rpm
are there other libraries I need ?

I found a reference on the web to a module called
libc6-pic.o, I don't see this anywhere in my libraries
is this what I need ?

Has anyone else seen this problem before and do they
know how to fix it ?

TIA

Wayne

__________________________________________________
Do You Yahoo!?
Find the one for you at Yahoo! Personals
http://personals.yahoo.com

Re: ld error " linking PIC files with non-PIC files "

[Ralf Baechle]

On Tue, Nov 13, 2001 at 12:09:48PM -0800, Wayne Gowcher wrote:

> I am trying to cross compile X using the redhat7.0
> distribution from the sgi mips site as my base. Most
> files compile OK, but they fail at the link stage with
> the following error :
> 
> linking PIC files with non-PIC files.
> 
> I created my cross-compile library files in
> /usr/mipsel-linux from the packages :
> glibc-2.2.2-1.mipsel.rpm &
> glibc-devel-2.2.2-1.mipsel.rpm
> are there other libraries I need ?
> 
> I found a reference on the web to a module called
> libc6-pic.o, I don't see this anywhere in my libraries
> is this what I need ?
> 
> Has anyone else seen this problem before and do they
> know how to fix it ?

Don't use non-pic code ever in userspace.  Actually it's very strange
that you hit this problem as gcc defaults to pic code so you should
try to find which of your object files or libraries are non-pic.

  Ralf

Re: ld error " linking PIC files with non-PIC files "

[Ralf Baechle]

On Mon, Nov 19, 2001 at 04:05:09PM +0800, renc stone wrote:

> That's the same problem as mine.
> I try to use 64bit long long div& mod in one module,
> and find I miss some _divdi3 and something like that.
> 
> when I try to link my module with libgcc.a, in my mipsel-glibc 2.95.3,
> the ld report the same thing.
> 
> Does it mean I can't use 64bit div in module? How to get rid of this error?

Again, it's a grave mistake to mix pic and non-pic libraries.

To solve this you must either supply your own non-pic versions of the
routines in question or - and better - try to avoid them.  In your case
take a look at <asm/div64.h> which supplies a 64-bit by 32-bit division
routine.

  Ralf